我应该使用哪个函数来估计R中的特定ARIMA模型?

时间:2016-07-26 20:06:57

标签: r time-series forecasting

我有一个大约200行的mydata.ts。我使用了静态测试,采取了差异并检查了PACFARIMA(1,1,1)(0,1,1)。所以我决定试试Arima。 我应该使用哪个R函数来查找拟合值和预测? arimaauto.arimasummary(model)? 我可以信任//Watcher for Bluetooth LE Services private void StartBLEWatcher() { int discoveredServices = 0; // Hook up handlers for the watcher events before starting the watcher OnBLEAdded = async (watcher, deviceInfo) => { Dispatcher.RunAsync(CoreDispatcherPriority.Low, async () => { Debug.WriteLine("OnBLEAdded: " + deviceInfo.Id); GattDeviceService service = await GattDeviceService.FromIdAsync(deviceInfo.Id); var services = service.GetAllIncludedServices(); int count0 = services.Count; //returns 0 Guid G = new Guid("0000ffe5-0000-1000-8000-00805f9b34fb"); var services2 = service.GetIncludedServices(G); int count = services2.Count; //returns 0 although this service "should" exist var characteristics = service.GetAllCharacteristics(); int count2 = characteristics.Count; //return 1 This is the Gatt service with Notify 上的MAPE,MAD和其他错误结果吗?因为我读了一个答案,并且说结果不是真实的,而是近似的或者是某种东西。

1 个答案:

答案 0 :(得分:0)

auto.arima将找到最佳的'最佳模型规范。基于AIC,BIC。

如果您知道订单(1,1,1)或(0,1,1),那么请使用预测包中的Arima(与arima相同,但更为一般)

Arima(your_data, order=c(1,1,1)) will give the basic answer.

forecast的文档。

然后可以使用预测功能完成实际的样本外预测。